Dashboard and Monitors
The signed-in app opens on the Dashboard. Next to it in the sidebar is Monitors, the list of every endpoint and domain, where each monitor opens in its card. Metrics charts how checks have gone over time. The bell and the Getting started checklist sit around them. For the record of what changed on a monitor, see History.
Dashboard
The Dashboard is the daily briefing: what needs a look now, when each certificate and registration turns Late, and whether alerts can be delivered. The Healthy, Open alerts, and Next expiry tiles, the Triggered card and the Runway load together, and the Alert delivery tile loads on its own. When the tiles and Triggered, or the Alert delivery tile, fail to load, they show an Update failed notice with a Retry button and keep their last good data while the rest carry on. The page refreshes itself every 60 seconds while it is open and visible, and the Refresh button refreshes it now. The line under the title says CertWatch is checking every hour, how long ago the newest check of an active monitor finished, by CertWatch's clock, so a wrong clock on your computer does not change it, and how many monitors it covers, with a pulsing green dot while checks run. It reads Checks delayed with an amber dot if CertWatch's checks fall behind, so you know how old the page is. Hover over it for the time of that check. Endpoints are checked every hour; a domain's page says how often its registration is checked.
Banners
Between the title and the tiles, the Dashboard shows at most these banners, in this order: a billing notice when a payment needs action, a cancellation is scheduled (with the date it takes effect), or your plan is enforced (with a link to each endpoint it paused), or Billing status unavailable when CertWatch cannot read it; CertWatch's checks are delayed since a time, while CertWatch's own checks are delayed, when each monitor keeps the status its last check gave; and a banner with an Upgrade link once you reach your plan's endpoint or domain limit. With none of these, there is no banner.
Tiles
Four tiles open the page. They count your active monitors only: paused monitors are left out.
Healthy gives the share of active monitors whose status is Healthy, such as "3 of 4", with one cell per monitor: Down first, then Urgent, Unreachable, and Late, then monitors in neither (a new monitor awaiting its first check, or a domain on an unsupported TLD) in a plain colour, then Healthy. Its chip counts the Unsupported domains, or else the monitors awaiting their first check, or, when every monitor is paused, the paused ones, or says "all Healthy". The caption counts your endpoints and domains.
Open alerts counts the monitors with an open alert and how many of those are not acknowledged. A domain's alert has no acknowledgement. A line shows how the count rose over the last 30 minutes, and the caption counts the open alerts by severity and says since when the oldest has been open, or reads "No open alerts". A chip counts the Down alerts when some open alerts are not Down.
Next expiry counts the days to the soonest certificate expiry among your Healthy endpoints, with its Late from date once CertWatch knows it, and marks each Healthy certificate that expires within 60 days on a 60-day line. A Triggered endpoint's certificate is in the Triggered list instead.
The fourth tile, Alert delivery, is described below. With nothing monitored yet, Add endpoint and Add domain take the place of the first three tiles, each opening Monitors' add form on that kind, and the Alert delivery tile stays beside them.
Triggered
Triggered lists the monitors you should look at: those whose status is Late, Urgent, Down, or Unreachable. A paused monitor, a monitor still within the time allowed for its first check, and a domain on a TLD with no registration data are not Triggered, so a new monitor does not appear here while its first check is on the way. If that check is overdue, the monitor is Triggered. Triggered is worked out from the status each time you open the page. It sends nothing; alerts follow their own rules (see Alerts).
While any monitor is Triggered, the owl at the top of the sidebar and the browser-tab icon show red eyes on every page with the sidebar, and the owl's tooltip names how many monitors are Triggered. They update every minute while the tab is visible, when you come back to the tab, when you move to another page, when you click Refresh on the Dashboard, and after you pause, resume, add or remove a monitor.
The card lists every Triggered monitor. Its title counts them, with the dot and the count in the colour of the worst status among them, and how many Down alerts are not acknowledged; the dot pulses while one is not. The rows sit in four groups, Down, then Urgent, then Unreachable, then Late, each headed by its count and, when any are, how many are acknowledged ("Late 2 · 1 acknowledged"). Inside a group, rows whose alert is not acknowledged come first; then a Late group puts the nearest expiry first, an Unreachable group the checks that began failing first, and the others the problem that started longest ago. A row you acknowledge or withdraw stays where it is until the page next updates. A line below the groups counts your Healthy monitors; with nothing Triggered, the card says "Nothing triggered" with that count.
Each row starts with a dot in its status's colour, pulsing on a Down row whose alert is not acknowledged, or in the acknowledged colour while an acknowledgement holds (hover over it for the status's name), then the monitor's name (an endpoint's label, else its host and port; a domain's name, tagged Domain). A second line gives the endpoint's host and port (left out when the name already is them) or the domain's registrar once the registry names one, the Reason, such as "Certificate problem" or "Host unreachable since Oct 3; last known Down 2 days ago" (a renewal that is simply due shows none, since the expiry date says it), and the expiry date once CertWatch has read one. An endpoint whose check failed keeps its last status until a second failed check, normally the follow-up about two minutes later, makes it Unreachable (an endpoint never read is Unreachable at once, and Urgent or Down outranks Unreachable), and its reason says the checks are failing: "Host unreachable since Oct 3". A muted endpoint carries a Muted badge. The filter box keeps the rows whose name, host, domain, or registrar contains what you type.
One row is selected, the first one until you pick another, and on a wide screen its detail fills the pane beside the list. Click a row, or press Enter on it, to select it. The selection stays on its row when the page updates or you filter, as long as the row is still listed, and moves to the first row once it is not. On a narrow screen there is no pane: a click or Enter opens the row's inline detail under it, and opening another closes it. The inline detail holds the same facts in short: what was found and what to check, an endpoint's renewal bar or a domain's expiry, the last check, one line for the alert's deliveries and the next alert, and the same actions.
The pane names the monitor and its status. For an endpoint it says since when its alert has been open, and who acknowledged it and when. Then come the alert's headline, every finding with its status, and the alert's How to fix steps, in a box in the status's colour. When the alert is about a single confirmed finding, the box gives only the addresses it was found on ("On 203.0.113.18"), since the headline already says it. When the host is unreachable or the checks are failing, the box also says so ("Host unreachable since Jun 14; last known Down 2 days ago"), and with no alert and no finding it gives the status's reason alone ("Certificate chain is invalid"). Next come the renewal points on a bar with today marked, each point labelled with its date and how far away it is ("in 4 days", "7 hours ago"), with the renewal mode, and the CA's suggested window when the dates come from it; what the last check saw, when it ran and when the next is due; and the alert's delivery. The renewal points' Change link opens the monitor card's Settings tab, where you set them. For an endpoint, the last check covers the addresses it tried (up to eight) that answered or failed: whether they all served the same certificate and when it expires, how many fail verification and why, or how many did not answer, with the first of them; Details, once the check has loaded, opens the monitor's Checks tab, which lists every address of every check. A domain's pane shows its expiry, registrar, transfer lock, DNSSEC, and when it was last checked. The alert's delivery says when the alert last went out ("Last sent 02:56"), with an Alert details link to the alert, and shows a tile for each channel the alert went to, and each other channel that can send (set up, in your plan, and for email verified), in Email, Slack, Webhook order: Sent, Failed, Retrying, Sending, Not sent when the alert should have gone to it and did not, or Turned off when your Alert channels do not send this severity to it. Its last line says what needs doing first: a failed delivery's reason with the link that fixes it ("Slack access rejected · Reconnect Slack"), a retry under way ("Retrying Slack at 12:05"), or what comes next: the next alert ("Next reminder 18:00", "Next: Urgent alert Jun 25, 06:00") and what holds it ("held by the mute until 18:00"), or, once you acknowledge, the message that still comes ("Acked · next: Urgent Jun 25, 06:00"). A date already past says how long ago, in hours on its first day ("13 hours ago"); hover over that for the exact time, in your time zone.
The pane's actions, the same on a monitor's inline detail apart from Retry:
- Acknowledge, as on the row; the message after it has Undo. Once you acknowledge, the pane shows Acked beside the status, and a note under the actions says who acknowledged it and when, such as "Acked by Taylor at 11:00.", with Undo at its end to withdraw it. The inline detail shows the same note, and its row shows Acked.
- Mute alerts, the same choice of lengths as on the monitor card.
- Check now, on every endpoint and domain. It reads "Checking…" until the result arrives, then a line under the actions reads "Last check 12:04 · Down", or "Last check 12:04 · Not conclusive · Host unreachable" when the check could not read the monitor; if the check takes longer than 30 seconds, the line says it is still running and the result shows at the next refresh. A monitor the check finds Healthy leaves the list.
- Pause monitoring, on a domain, after you confirm; the message after it has Undo.
- Open monitor page, which opens the monitor's card.
- Retry at the end of the delivery's last line, on the first failed delivery when it can be retried, as on the alert's page; the pane only. The alert's page retries each channel.
When an action fails, its reason shows under the actions; a delivery's Retry reports under the delivery's last line, and Mute, and the Undo in the message after an Acknowledge or a pause, report a failure in a message. A domain has no Acknowledge or Mute.
An endpoint row whose alert is not acknowledged has Acknowledge, and the message after it has Undo. Once acknowledged it shows Acked, with Undo to withdraw it; hover over Acked for what it holds (see Acknowledge). While two or more Down endpoints have an alert that is not acknowledged, Acknowledge all acknowledges each of them after you confirm, and one message says how many it acknowledged, with Undo to withdraw each of them. A domain's alert has no acknowledgement.
On the keyboard, / goes to the filter, where ↓ and ↑ move the highlight through the listed rows and Enter opens the highlighted row; j and k move the selection down and up the list, and Enter selects the row (on a narrow screen it opens or closes its inline detail); Esc closes what is open (the shortcuts list, then a confirm, then the Alert delivery channels, then the inline detail), else clears the filter; and ? lists the shortcuts. Apart from Esc, none of them act while you type in a text field, and none act while you hold Ctrl, Alt, or Cmd, except two that work on every page: ⌘K (Ctrl K), even in a field, opens Jump to, which finds a monitor and opens it, or goes to a page; and outside a text field, ⌘Z (Ctrl Z) undoes the latest acknowledgement or pause while its message shows.
Runway
The Runway puts every active monitor on one timeline, from a week before today to the end of the range you choose: 90 days, 6 months, or 1 year. All, Certificates, and Domains narrow it to one kind. Each lane names the monitor, a link to its card, with the endpoint's host and port or "Domain" and the domain's registrar, and draws its track: Healthy, Late, and Urgent segments split at the dates CertWatch works out for it, ending at the expiry. An endpoint's lane ends when the first certificate in its chain expires, which can be an intermediate before the certificate itself. A domain turns Late 30 days and Urgent 7 days before its registration expires. A dot marks an alert that went out. It pulses until an endpoint's alert is acknowledged, unless your system asks for reduced motion.
A Down monitor whose certificate or registration has expired shows Down up to today, with when it expired ("Expired Nov 3 · 13 hours ago"; hover over how long ago for the exact time). A monitor that is Down for another reason, such as a broken chain, shows Down from today to its expiry, with the reason. An Unreachable monitor shows a dashed line from today with its reason, because CertWatch cannot read its dates. An endpoint whose alert is acknowledged shows Acked, with Undo to withdraw the acknowledgement.
The lanes sit in sections. Triggered comes first, worst first: Down, Urgent, Unreachable, then Late. Certificates and Domains hold the Healthy ones that turn Late within the range, soonest first. Each section's count is beside its title. When more turn Late after the range, a line says how many and which is next, with Show to list them too; when none falls within the range, its button widens the range to the one that holds the next, or reads Show when even a year does not reach it. Each section shows 25 lanes, then Show 25 more. A last line counts the monitors with no expiry date yet, such as one awaiting its first check or a domain on a TLD with no registration data. Paused monitors are left out.
Hover over a lane's track, or tap it on a phone, to see the Late from, Urgent from and Expires dates it has; an Unreachable lane has none. The info icon at the end of the legend says the same. On a phone, tap the track again, tap elsewhere, or scroll to close them.
Alert delivery
This tile shows whether alerts can reach you. It lists the alert channels you have set up, except an alert email you have not verified: the banner at the top of the app, with its Resend email button, is the one prompt for that. A channel is failing when it has current delivery failures, when it is not usable or not in your plan, or when no alert has gone out today and its last test send failed. Any other channel is ready, including one that is Not yet tested: its current destination has never had an accepted alert or test send, which is neutral rather than a failure. An email address whose latest accepted alert or test email later bounced or was suppressed also returns to Not yet tested, and replacing an email address or a URL starts it over.
Until an alert goes out today, the tile reads Ready or how many channels are ready. Once alerts have gone out since midnight in your time zone, it reads how many of them were delivered. Fix delivery marks a failing channel, and the lines under the tile name each one. With no usable channel, it says "No channel can deliver an alert."
Select the tile, or press Enter or Space on it, to open its channels below the row. Each channel's row gives today's accepted and failed deliveries ("Today: 3 accepted, 1 failed"), else its last test, else Not yet tested. Send test alert sends a test to each usable channel and reports each result on that channel's row. A channel with current delivery failures lists each with a Details link to the alert's page, and Retry ("Retry 5 failed") sends every failed alert on that channel from the last 90 days again, to its current destination, and its count includes them all; the row says how many are retrying until the page next refreshes. Failures from an earlier destination, and those to an unverified alert email, are listed on Alert history and the alert's page rather than here. When no channel is usable, a row says "No channel can deliver an alert." and links to the Alerting page, unless the only channel set up is an unverified alert email. Add a channel names the channels your plan includes that you have not set up, with an Add button to Alerting.
When a channel is confirmed failing and another usable destination exists, CertWatch also tells you there (a Delivery notice, described on Alerts). With no other usable destination, no notice is sent and this tile is the signal.
Monitors
Monitors lists every endpoint and domain on the account, paused ones included. The All, Endpoints, and Domains buttons, each with its count, narrow the list to one kind, and the choice stays in the address, so a reload or a shared link shows the same kind. The filter box keeps the monitors whose name, host, or registrar contains what you type. Each row gives the monitor's name (an endpoint's label, else its host and port; a domain's name, tagged Domain) with the host and port or the registrar under it, its status and reason (an Unsupported domain's reason is in its status's tooltip), its expiry date, when it was last checked, and its alert state and actions. Sort by Monitor, Status, Expires, or Last check from the column headers; the list shows 50 rows, then Show more, and a footer counts the rows shown while a filter or the paging hides some.
An endpoint with an open alert has Acknowledge in its alert state, and once acknowledged Acked and Undo; the message after it says "Acknowledged" and the name. Hover over a row, or move to it with the keyboard, to show its actions: the mute menu on an endpoint, Pause or Resume, and a menu with Open, Edit, Check now, and Remove, which asks "Remove <name>?" first: removing deletes the monitor's checks and certificates, and its History stays for 90 days. A muted endpoint shows a mark while its actions are hidden.
On the keyboard, / goes to the filter, j and k move down and up the list, Enter opens the monitor, Esc closes what is open or clears the filter, and ? lists the shortcuts.
Adding a monitor
Add opens the form above the list, or moves to it while it is open; Cancel or Esc closes it. Choose Endpoint and give a hostname, a port (443 unless you change it), and an optional label, or choose Domain and give a domain name. Test (an endpoint) or Look up (a domain) reads the monitor before you add it: an endpoint's certificate status and expiry, or a domain's registrar and expiry. At your plan's limit the form says so once, such as "Endpoint limit reached (4 of 4)", with an Upgrade link. A domain whose first check finds its TLD unsupported shows the message "Added example.io as Unsupported" with Learn more. Once added, the monitor's first check runs at once and its row reads Pending until the result arrives. The Dashboard's empty state and the Getting started checklist open Monitors with this form already open. See Setup for what to enter.
The monitor card
Selecting a monitor opens its card over the list. The card's address names the monitor and its tab, such as
/monitors/<id>/settings, so a link from an alert opens the card on the right tab, and
links CertWatch sent before the card existed still open it. The card keeps one size on every tab; a tab
with more than fits scrolls inside the card. The header names the monitor with its status and
holds its actions: Acknowledge or Undo and the mute menu on an endpoint; Pause or
Resume, Check now, and Remove. The close button, a click outside the card, and
Esc go back to the list as you left it.
- Overview: why the monitor has its status, with each finding; on an endpoint, Alert state while it has something to state: the open alert and its next message, who acknowledged it and when, and what the mute on a paused endpoint holds back; an endpoint's renewal points, with a Change link to its Settings tab, and what the check that read its certificate found: its common name, chain, subject alternative names, serial, issuer (the CA's name with its common name, such as "Let's Encrypt (YE2)"), and the addresses it reached, with a mark and a hover explanation beside one that failed or was not tried; a domain's registration points, where it was looked up, and registry signals.
- History: the monitor's History.
- Checks: an endpoint's graph (Over time) and the monitor's recent checks (Each check), where checks in a row that found the same certificate show as one row, such as "12 identical checks", and a known chain error reads in plain words. A check says why it ran when CertWatch recorded it: Scheduled, Check now, Added, Resumed, Renewal setting changed, or Confirmation (the check a couple of minutes after one that first saw a finding, which looks for it again: see Findings); some older checks have no reason. Open a row of identical checks to see each one's.
- Settings: an endpoint's label, its hostname and port, which cannot be changed, then its renewal mode: how the certificate is renewed, with each mode's Late and Urgent dates before you save (see Renewal mode); a domain's name, which cannot be changed.
A link to a monitor that was removed, or that is on another account, opens Monitors with the message "That monitor isn't on this account."
Metrics
Metrics is not in the sidebar: open it at https://app.certwatch.sh/app/metrics. It charts check results across every endpoint and domain on the account: for each period, how many checks
ended Healthy, Late, Urgent, Down, or Unreachable. A check of a domain on a TLD with no registration data counts
in none of them. Choose 7 days (hourly points), 30 days, or 90 days (daily points). The chart can
only show checks you still have: your plan keeps check results for 30, 90, or 365 days (see
Monitoring behavior), so a Free account's 90-day range shows its last
30 days. A new account shows "No checks yet" until its first checks finish. For one monitor's own checks, open
its card's Checks tab (an endpoint's also has a graph).
The bell
The bell in the header lists Notifications: to-do items worked out from your account as it is now. At present they are usage warnings, from 80 percent of your plan's endpoint or domain limit ("Approaching endpoint limit", or "Endpoint limit reached" at 100 percent, and the same for domains), each with a View plans link. The number on the bell counts them, and with none it reads "You're all caught up".
A Notification is not stored and cannot be dismissed: it disappears when the account no longer matches, for example when you remove an endpoint or upgrade. Four things that look similar are not in the bell. Alerts about a monitor go to your alert channels and are recorded in Alert history. Delivery failures appear on the Dashboard's Alert delivery tile. An unverified alert email has its own banner. Usage Notices (the emails about reaching your limits) are described on Alerts.
Getting started checklist
New accounts see a Getting started checklist at the bottom of the sidebar, with a progress bar. Selecting a step takes you to the page where you finish it.
| Step | Done when | Opens |
|---|---|---|
| Add an endpoint | The account has at least one endpoint. | Monitors, with the add form open on Endpoint |
| Confirm your first check | An endpoint has completed a check. | Monitors, showing endpoints |
| Verify your alert email | You have used the confirmation link CertWatch emailed you. | Settings |
| Send a test alert | A test send on Alerting or the Dashboard succeeds for email, Slack, or webhook. | Alerting |
The checklist disappears once every step is done, or when you dismiss it with the close button. Whether you dismissed it, and whether a test alert was sent, are remembered in the browser you used, so the checklist can come back in another browser or after clearing site data; the first, second, and third steps follow your account, not the browser. The Setup page walks through the same steps in prose.